Increasing evidence indicates that submicroscopic genomic imbalances (deletions, duplications) play a major role in human disease, especially neurodevelopmental disorders such as mental retardation and autism. We propose to test the hypothesis that subtelomeric and pericentromeric regions of the human genome are highly prone to genomic instability and cause a disproportionate percentage of developmental disabilities due to their unique genomic topology. Their complex, mosaic structures are rich in repetitive and duplicated sequences, which have been recalcitrant to whole genome sequencing efforts. Targeted approaches are needed to elucidate their genomic architecture, evolution and role in human disease. We propose the following Specific Aims: 1) Investigate the genomic architecture of human subtelomeric and pericentromeric regions. We will establish the boundary between the unique and repetitive DNA zones for each subtelomeric and pericentromeric region and create "molecular rulers" up to 5 Mb from each telomere or centromere. Each region will be classified (complex vs. simple; polymorphic vs. non-polymorphic) to test the hypothesis that a complex or polymorphic structure may increase instability. 2) Determine the frequency, patterns and mechanisms of chromosome abnormalities at subtelomeric and pericentromeric regions. Patients with abnormalities involving subtelomeric or pericentromeric regions will be analyzed by FISH and/or array CGH using our novel molecular ruler reagents. Gene dosage effects will be assessed in a) MR or autistic individuals and b) phenotypically normal individuals with subtelomeric or pericentromeric imbalance. 3) Develop a "human gene dosage map" for the pericentromeric and subtelomeric regions of human chromosomes using 2 approaches: a) Computational studies of gene content within the precisely calibrated segments of dosage imbalance will be used as a high-resolution approach to genotype-phenotype correlation in humans, b) Functional studies of gene expression changes associated with genomic imbalance will be globally surveyed using oligonucleotide arrays of ~33,000 genes followed by targeted analyses using RT-PCR and/or northern analyses to assess expression of genes directly involved in the chromosomal event as well as neighboring genes (position effects) and a genome wide survey (gene interaction effects).